mathmodel-skill
A skill that walks a student team through a math modeling contest in ten stages, from choosing a problem to submission
- 295 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- The agent runs the project's Python scripts and writes files into the workspace, including the state log
- Building the paper calls outside programs, Pandoc and LaTeX
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/mathmodel-skillDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The agent runs the project's Python scripts and writes files into the workspace, including the state log
- Building the paper calls outside programs, Pandoc and LaTeX
This is third-party code. Review the repository files before installing.
What it does
mathmodel-skill takes a team through a math modeling contest in ten stages, from problem choice and breakdown to paper writing and a final review. It supports three contests: China's CUMCM, the international MCM/ICM and the Chinese Diangong Cup. The agent asks numbered questions and does the technical steps itself. Decisions, scores and rollbacks live in state/decision_log.json. Work can therefore continue in a new chat, with another teammate or in another agent. A script, score_artifact.py, recomputes each stage verdict, and failing one criterion cannot hide behind an average. If a sub-question scores below the threshold only that one is reworked and the other results stay as they were. Before submission the skill checks five gates: the current contest rules, anonymity, page count, AI-use disclosure and supporting materials. Placeholders and misplaced template markers block the ready mark. Papers compile with XeLaTeX (CUMCM and the Diangong Cup) and pdfLaTeX (MCM/ICM). The phrase "show progress" at any stage prints the state, a countdown to the deadline and one next step.
Who it is for. For student teams in math modeling contests who work in Codex or Claude Code.
Good fit when
- The team prepares for CUMCM, MCM/ICM or the Diangong Cup and has 72–96 hours
- Decisions must survive a change of teammate, chat or agent: the state log lives in the project
- You want to see which sub-question lags in score and rework only that one
- Before submission you need to check anonymity, page count and AI-use disclosure against this year's rules
- You need a quick progress summary with a deadline countdown
Not a fit when
- The contest is different: the skill knows only the three listed competitions
- You need one finished paper with no teamwork and no state log
- Your contest rules forbid AI assistance, and that must be settled first
- The machine has no Python 3.10 or newer for the helper scripts
Example request
Start modeling for CUMCM: the problem is not published yet, 72 hours until submissionLimitations
- Documentation and the main skill text are in Chinese. The agent replies in the team's language, but templates and contest rules stay Chinese and English.
- Building the final PDF needs Pandoc and TeX Live or MiKTeX. The built-in simple converter only checks structure.
- Numerical computing dependencies are installed separately with pip from templates/shared/requirements.txt.
- Contest rules change every year, so the current regulation must be checked against the official one.
How to disable. Delete the mathmodel-skill folder from the skills directory (~/.claude/skills or ~/.agents/skills). The project folder with the state log stays with you.
FAQ
How do I install mathmodel-skill in Claude Code?
Run git clone https://github.com/handsomeZR-netizen/mathmodel-skill.git ~/.claude/skills/mathmodel-skill, create a project folder, start claude there and say: start modeling.
What does the show progress command do?
The status.py script prints verdicts for the ten stages, sub-question state and the five gates. It also shows a deadline countdown, a mode recommendation and one next step.
Can I start in Codex and finish in Claude Code?
Yes. State lives in state/decision_log.json inside the project, so both agents continue from the same checkpoint.
Which contests are supported?
China's CUMCM, the international MCM/ICM and the Chinese Diangong Cup. For each, the competitions folder holds rules, abstract templates and notes on typical mistakes.
What blocks submission?
Placeholders in the text, missing sections and misplaced template markers. The ready mark cannot be set until the check gates pass.
Discussion
No questions or reviews yet.
Related
SkillsA catalog of 28 mental-model skills for agents: diagnosis, decisions, systems, risk, strategy and idea generation
A desktop app and skm CLI: one copy of each skill in a central hub, symlinked into every AI tool
The author's personal skills for reading papers and books, dissecting concepts, writing and PNG cards; instructions in Chinese, installed via skills CLI
Qiushi Skill
求是 Skill
A set of methodology skills: investigate the facts first, find the main contradiction, validate in practice and finish the job
A skills pack for founders and solo operators: decisions, research, second brain, content rotation, financial cadence
A skill that builds an anonymized developer profile and a vertical poster from local Claude Code and other CLI history
README in short
The repository lists typical contest failures: a model changed while the abstract was not, assumptions kept only in chat, format violations found right before submission. It then covers the ten-stage flow, command-line tools, the three contests and installation for Codex and Claude Code. The author stresses that the skill does not write a paper in one prompt or decide for the team.
README badge
Original SKILL.md
--- name: mathmodel-skill description: CUMCM 国赛、MCM/ICM 美赛与电工杯数学建模竞赛的端到端协作工作流。Use when a user explicitly works on one of these modeling contests or asks to run/review a modeling-competition paper from problem selection through modeling, solving, robustness, writing, compliance, and final submission review. Provides 10 stages, persistent decision state, competition-specific rules/templates, deterministic scoring helpers, numbered decisions, and Codex/Claude Code handoff. Do not trigger for generic model selection, ordinary data analysis, or non-competition paper review. --- # mathmodel-skill — 数学建模三竞赛工作流 (v6.2) 10 阶段把 72–96 小时的竞赛协作变成可恢复、可检查的流程。用户回答关键问题，agent 维护状态与脚本。每阶段产出经过 rubric 自评、定向精修与跨阶段一致性回检；Stage 8–9 先遵守当届官方规则，再做多视角终审。CUMCM 包含 91 份来源文档，其中 59 份进入文本统计；MCM/电工杯经验统计明确为 `n=0`，不提供合成分位。 **v6.2 更新**: 新增 `scripts/init_workspace.py`（只创建、不覆盖的工作区初始化）与 `scripts/status.py`（只读进度看板：阶段 verdict、per-Qi、合规门、截止倒计时、模式建议与下一步）；启动与“看进度”改由脚本确定性完成。v6.1 的规则基线、AI 披露链路、fail-closed 模板与 doctor 预检保持不变。 --- ## Codex 原生入口 Codex 优先按 skill 目录发现本文件: - 用户级安装: `$HOME/.agents/skills/mathmodel-skill/` - 项目级安装: `<repo>/.agents/skills/mathmodel-skill/` - UI 元数据: `agents/openai.yaml` - 插件分发元数据: `.codex-plugin/plugin.json` + `skills/mathmodel-skill/SKILL.md` shim - 项目指导: `AGENTS.md` 仍可作为 repo / workspace 级 instructions, 但不是唯一入口 当 skill 已安装后, 用户可直接说"开始建模"或显式说"使用 `$mathmodel-skill` 开始建模"。 --- ## Harness 兼容 (Claude Code / Codex) 本 skill v6.2 以 Codex Skills 为一等入口, 同时保持 harness-agnostic 设计: | harness | 入口文件 | 用户交互工具 | 状态文件 | |---------|---------|-------------|---------| | Claude Code | `SKILL.md` (本文件) | `AskUserQuestion` 工具 | `<cwd>/state/decision_log.json` | | Codex CLI / Codex app | skill 目录中的 `SKILL.md` + 可选 `AGENTS.md` | markdown 编号列表 | 同上 (**互通**) | 跨 harness 互通: day 1 用 Codex 跑 stage 0-2, day 2 切回 Claude Code 接着 stage 3+, 状态完全保留。详见 `references/harness_compat.md`。 ---